3 perlop - Perl operators and precedence
7 Perl operators have the following associativity and precedence,
8 listed from highest precedence to lowest. Operators borrowed from
9 C keep the same precedence relationship with each other, even where
10 C's precedence is slightly screwy. (This makes learning Perl easier
11 for C folks.) With very few exceptions, these all operate on scalar
12 values only, not array values.
14 left terms and list operators (leftward)
18 right ! ~ \ and unary + and -
23 nonassoc named unary operators
24 nonassoc < > <= >= lt gt le ge
25 nonassoc == != <=> eq ne cmp
34 nonassoc list operators (rightward)
39 In the following sections, these operators are covered in precedence order.
41 Many operators can be overloaded for objects. See L<overload>.
45 =head2 Terms and List Operators (Leftward)
47 A TERM has the highest precedence in Perl. They include variables,
48 quote and quote-like operators, any expression in parentheses,
49 and any function whose arguments are parenthesized. Actually, there
50 aren't really functions in this sense, just list operators and unary
51 operators behaving as functions because you put parentheses around
52 the arguments. These are all documented in L<perlfunc>.
54 If any list operator (print(), etc.) or any unary operator (chdir(), etc.)
55 is followed by a left parenthesis as the next token, the operator and
56 arguments within parentheses are taken to be of highest precedence,
57 just like a normal function call.
59 In the absence of parentheses, the precedence of list operators such as
60 C<print>, C<sort>, or C<chmod> is either very high or very low depending on
61 whether you are looking at the left side or the right side of the operator.
64 @ary = (1, 3, sort 4, 2);
65 print @ary; # prints 1324
67 the commas on the right of the sort are evaluated before the sort,
68 but the commas on the left are evaluated after. In other words,
69 list operators tend to gobble up all arguments that follow, and
70 then act like a simple TERM with regard to the preceding expression.
71 Be careful with parentheses:
73 # These evaluate exit before doing the print:
74 print($foo, exit); # Obviously not what you want.
75 print $foo, exit; # Nor is this.
77 # These do the print before evaluating exit:
78 (print $foo), exit; # This is what you want.
79 print($foo), exit; # Or this.
80 print ($foo), exit; # Or even this.
84 print ($foo & 255) + 1, "\n";
86 probably doesn't do what you expect at first glance. See
87 L<Named Unary Operators> for more discussion of this.
89 Also parsed as terms are the C<do {}> and C<eval {}> constructs, as
90 well as subroutine and method calls, and the anonymous
91 constructors C<[]> and C<{}>.
93 See also L<Quote and Quote-like Operators> toward the end of this section,
94 as well as L<"I/O Operators">.
96 =head2 The Arrow Operator
98 "C<-E<gt>>" is an infix dereference operator, just as it is in C
99 and C++. If the right side is either a C<[...]>, C<{...}>, or a
100 C<(...)> subscript, then the left side must be either a hard or
101 symbolic reference to an array, a hash, or a subroutine respectively.
102 (Or technically speaking, a location capable of holding a hard
103 reference, if it's an array or hash reference being used for
104 assignment.) See L<perlreftut> and L<perlref>.
106 Otherwise, the right side is a method name or a simple scalar
107 variable containing either the method name or a subroutine reference,
108 and the left side must be either an object (a blessed reference)
109 or a class name (that is, a package name). See L<perlobj>.
111 =head2 Auto-increment and Auto-decrement
113 "++" and "--" work as in C. That is, if placed before a variable, they
114 increment or decrement the variable before returning the value, and if
115 placed after, increment or decrement the variable after returning the value.
117 The auto-increment operator has a little extra builtin magic to it. If
118 you increment a variable that is numeric, or that has ever been used in
119 a numeric context, you get a normal increment. If, however, the
120 variable has been used in only string contexts since it was set, and
121 has a value that is not the empty string and matches the pattern
122 C</^[a-zA-Z]*[0-9]*$/>, the increment is done as a string, preserving each
123 character within its range, with carry:
125 print ++($foo = '99'); # prints '100'
126 print ++($foo = 'a0'); # prints 'a1'
127 print ++($foo = 'Az'); # prints 'Ba'
128 print ++($foo = 'zz'); # prints 'aaa'
130 The auto-decrement operator is not magical.
132 =head2 Exponentiation
134 Binary "**" is the exponentiation operator. It binds even more
135 tightly than unary minus, so -2**4 is -(2**4), not (-2)**4. (This is
136 implemented using C's pow(3) function, which actually works on doubles
139 =head2 Symbolic Unary Operators
141 Unary "!" performs logical negation, i.e., "not". See also C<not> for a lower
142 precedence version of this.
144 Unary "-" performs arithmetic negation if the operand is numeric. If
145 the operand is an identifier, a string consisting of a minus sign
146 concatenated with the identifier is returned. Otherwise, if the string
147 starts with a plus or minus, a string starting with the opposite sign
148 is returned. One effect of these rules is that C<-bareword> is equivalent
151 Unary "~" performs bitwise negation, i.e., 1's complement. For example,
152 C<0666 &~ 027> is 0640. (See also L<Integer Arithmetic> and L<Bitwise
155 Unary "+" has no effect whatsoever, even on strings. It is useful
156 syntactically for separating a function name from a parenthesized expression
157 that would otherwise be interpreted as the complete list of function
158 arguments. (See examples above under L<Terms and List Operators (Leftward)>.)
160 Unary "\" creates a reference to whatever follows it. See L<perlreftut>
161 and L<perlref>. Do not confuse this behavior with the behavior of
162 backslash within a string, although both forms do convey the notion
163 of protecting the next thing from interpolation.
165 =head2 Binding Operators
167 Binary "=~" binds a scalar expression to a pattern match. Certain operations
168 search or modify the string $_ by default. This operator makes that kind
169 of operation work on some other string. The right argument is a search
170 pattern, substitution, or transliteration. The left argument is what is
171 supposed to be searched, substituted, or transliterated instead of the default
172 $_. The return value indicates the success of the operation. (If the
173 right argument is an expression rather than a search pattern,
174 substitution, or transliteration, it is interpreted as a search pattern at run
175 time. This can be is less efficient than an explicit search, because the
176 pattern must be compiled every time the expression is evaluated.
178 Binary "!~" is just like "=~" except the return value is negated in
181 =head2 Multiplicative Operators
183 Binary "*" multiplies two numbers.
185 Binary "/" divides two numbers.
187 Binary "%" computes the modulus of two numbers. Given integer
188 operands C<$a> and C<$b>: If C<$b> is positive, then C<$a % $b> is
189 C<$a> minus the largest multiple of C<$b> that is not greater than
190 C<$a>. If C<$b> is negative, then C<$a % $b> is C<$a> minus the
191 smallest multiple of C<$b> that is not less than C<$a> (i.e. the
192 result will be less than or equal to zero).
193 Note than when C<use integer> is in scope, "%" give you direct access
194 to the modulus operator as implemented by your C compiler. This
195 operator is not as well defined for negative operands, but it will
198 Binary "x" is the repetition operator. In scalar context, it
199 returns a string consisting of the left operand repeated the number of
200 times specified by the right operand. In list context, if the left
201 operand is a list in parentheses, it repeats the list.
203 print '-' x 80; # print row of dashes
205 print "\t" x ($tab/8), ' ' x ($tab%8); # tab over
207 @ones = (1) x 80; # a list of 80 1's
208 @ones = (5) x @ones; # set all elements to 5
211 =head2 Additive Operators
213 Binary "+" returns the sum of two numbers.
215 Binary "-" returns the difference of two numbers.
217 Binary "." concatenates two strings.
219 =head2 Shift Operators
221 Binary "<<" returns the value of its left argument shifted left by the
222 number of bits specified by the right argument. Arguments should be
223 integers. (See also L<Integer Arithmetic>.) Shifting more than the
224 width of an integer in bits (usually 32 or 64) produces undefined
225 (platform dependent) results.
227 Binary ">>" returns the value of its left argument shifted right by
228 the number of bits specified by the right argument. Arguments should
229 be integers. (See also L<Integer Arithmetic>.) Shifting more than
230 the width of an integer in bits (usually 32 or 64) produces undefined
231 (platform dependent) results.
233 =head2 Named Unary Operators
235 The various named unary operators are treated as functions with one
236 argument, with optional parentheses. These include the filetest
237 operators, like C<-f>, C<-M>, etc. See L<perlfunc>.
239 If any list operator (print(), etc.) or any unary operator (chdir(), etc.)
240 is followed by a left parenthesis as the next token, the operator and
241 arguments within parentheses are taken to be of highest precedence,
242 just like a normal function call. Examples:
244 chdir $foo || die; # (chdir $foo) || die
245 chdir($foo) || die; # (chdir $foo) || die
246 chdir ($foo) || die; # (chdir $foo) || die
247 chdir +($foo) || die; # (chdir $foo) || die
249 but, because * is higher precedence than ||:
251 chdir $foo * 20; # chdir ($foo * 20)
252 chdir($foo) * 20; # (chdir $foo) * 20
253 chdir ($foo) * 20; # (chdir $foo) * 20
254 chdir +($foo) * 20; # chdir ($foo * 20)
256 rand 10 * 20; # rand (10 * 20)
257 rand(10) * 20; # (rand 10) * 20
258 rand (10) * 20; # (rand 10) * 20
259 rand +(10) * 20; # rand (10 * 20)
261 See also L<"Terms and List Operators (Leftward)">.
263 =head2 Relational Operators
265 Binary "E<lt>" returns true if the left argument is numerically less than
268 Binary "E<gt>" returns true if the left argument is numerically greater
269 than the right argument.
271 Binary "E<lt>=" returns true if the left argument is numerically less than
272 or equal to the right argument.
274 Binary "E<gt>=" returns true if the left argument is numerically greater
275 than or equal to the right argument.
277 Binary "lt" returns true if the left argument is stringwise less than
280 Binary "gt" returns true if the left argument is stringwise greater
281 than the right argument.
283 Binary "le" returns true if the left argument is stringwise less than
284 or equal to the right argument.
286 Binary "ge" returns true if the left argument is stringwise greater
287 than or equal to the right argument.
289 =head2 Equality Operators
291 Binary "==" returns true if the left argument is numerically equal to
294 Binary "!=" returns true if the left argument is numerically not equal
295 to the right argument.
297 Binary "E<lt>=E<gt>" returns -1, 0, or 1 depending on whether the left
298 argument is numerically less than, equal to, or greater than the right
301 Binary "eq" returns true if the left argument is stringwise equal to
304 Binary "ne" returns true if the left argument is stringwise not equal
305 to the right argument.
307 Binary "cmp" returns -1, 0, or 1 depending on whether the left argument is stringwise
308 less than, equal to, or greater than the right argument.
310 "lt", "le", "ge", "gt" and "cmp" use the collation (sort) order specified
311 by the current locale if C<use locale> is in effect. See L<perllocale>.
315 Binary "&" returns its operators ANDed together bit by bit.
316 (See also L<Integer Arithmetic> and L<Bitwise String Operators>.)
318 =head2 Bitwise Or and Exclusive Or
320 Binary "|" returns its operators ORed together bit by bit.
321 (See also L<Integer Arithmetic> and L<Bitwise String Operators>.)
323 Binary "^" returns its operators XORed together bit by bit.
324 (See also L<Integer Arithmetic> and L<Bitwise String Operators>.)
326 =head2 C-style Logical And
328 Binary "&&" performs a short-circuit logical AND operation. That is,
329 if the left operand is false, the right operand is not even evaluated.
330 Scalar or list context propagates down to the right operand if it
333 =head2 C-style Logical Or
335 Binary "||" performs a short-circuit logical OR operation. That is,
336 if the left operand is true, the right operand is not even evaluated.
337 Scalar or list context propagates down to the right operand if it
340 The C<||> and C<&&> operators differ from C's in that, rather than returning
341 0 or 1, they return the last value evaluated. Thus, a reasonably portable
342 way to find out the home directory (assuming it's not "0") might be:
344 $home = $ENV{'HOME'} || $ENV{'LOGDIR'} ||
345 (getpwuid($<))[7] || die "You're homeless!\n";
347 In particular, this means that you shouldn't use this
348 for selecting between two aggregates for assignment:
350 @a = @b || @c; # this is wrong
351 @a = scalar(@b) || @c; # really meant this
352 @a = @b ? @b : @c; # this works fine, though
354 As more readable alternatives to C<&&> and C<||> when used for
355 control flow, Perl provides C<and> and C<or> operators (see below).
356 The short-circuit behavior is identical. The precedence of "and" and
357 "or" is much lower, however, so that you can safely use them after a
358 list operator without the need for parentheses:
360 unlink "alpha", "beta", "gamma"
361 or gripe(), next LINE;
363 With the C-style operators that would have been written like this:
365 unlink("alpha", "beta", "gamma")
366 || (gripe(), next LINE);
368 Use "or" for assignment is unlikely to do what you want; see below.
370 =head2 Range Operators
372 Binary ".." is the range operator, which is really two different
373 operators depending on the context. In list context, it returns an
374 array of values counting (up by ones) from the left value to the right
375 value. If the left value is greater than the right value then it
376 returns the empty array. The range operator is useful for writing
377 C<foreach (1..10)> loops and for doing slice operations on arrays. In
378 the current implementation, no temporary array is created when the
379 range operator is used as the expression in C<foreach> loops, but older
380 versions of Perl might burn a lot of memory when you write something
383 for (1 .. 1_000_000) {
387 In scalar context, ".." returns a boolean value. The operator is
388 bistable, like a flip-flop, and emulates the line-range (comma) operator
389 of B<sed>, B<awk>, and various editors. Each ".." operator maintains its
390 own boolean state. It is false as long as its left operand is false.
391 Once the left operand is true, the range operator stays true until the
392 right operand is true, I<AFTER> which the range operator becomes false
393 again. It doesn't become false till the next time the range operator is
394 evaluated. It can test the right operand and become false on the same
395 evaluation it became true (as in B<awk>), but it still returns true once.
396 If you don't want it to test the right operand till the next
397 evaluation, as in B<sed>, just use three dots ("...") instead of
398 two. In all other regards, "..." behaves just like ".." does.
400 The right operand is not evaluated while the operator is in the
401 "false" state, and the left operand is not evaluated while the
402 operator is in the "true" state. The precedence is a little lower
403 than || and &&. The value returned is either the empty string for
404 false, or a sequence number (beginning with 1) for true. The
405 sequence number is reset for each range encountered. The final
406 sequence number in a range has the string "E0" appended to it, which
407 doesn't affect its numeric value, but gives you something to search
408 for if you want to exclude the endpoint. You can exclude the
409 beginning point by waiting for the sequence number to be greater
410 than 1. If either operand of scalar ".." is a constant expression,
411 that operand is implicitly compared to the C<$.> variable, the
412 current line number. Examples:
414 As a scalar operator:
416 if (101 .. 200) { print; } # print 2nd hundred lines
417 next line if (1 .. /^$/); # skip header lines
418 s/^/> / if (/^$/ .. eof()); # quote body
420 # parse mail messages
422 $in_header = 1 .. /^$/;
423 $in_body = /^$/ .. eof();
424 # do something based on those
426 close ARGV if eof; # reset $. each file
431 for (101 .. 200) { print; } # print $_ 100 times
432 @foo = @foo[0 .. $#foo]; # an expensive no-op
433 @foo = @foo[$#foo-4 .. $#foo]; # slice last 5 items
435 The range operator (in list context) makes use of the magical
436 auto-increment algorithm if the operands are strings. You
439 @alphabet = ('A' .. 'Z');
441 to get all normal letters of the alphabet, or
443 $hexdigit = (0 .. 9, 'a' .. 'f')[$num & 15];
445 to get a hexadecimal digit, or
447 @z2 = ('01' .. '31'); print $z2[$mday];
449 to get dates with leading zeros. If the final value specified is not
450 in the sequence that the magical increment would produce, the sequence
451 goes until the next value would be longer than the final value
454 =head2 Conditional Operator
456 Ternary "?:" is the conditional operator, just as in C. It works much
457 like an if-then-else. If the argument before the ? is true, the
458 argument before the : is returned, otherwise the argument after the :
459 is returned. For example:
461 printf "I have %d dog%s.\n", $n,
462 ($n == 1) ? '' : "s";
464 Scalar or list context propagates downward into the 2nd
465 or 3rd argument, whichever is selected.
467 $a = $ok ? $b : $c; # get a scalar
468 @a = $ok ? @b : @c; # get an array
469 $a = $ok ? @b : @c; # oops, that's just a count!
471 The operator may be assigned to if both the 2nd and 3rd arguments are
472 legal lvalues (meaning that you can assign to them):
474 ($a_or_b ? $a : $b) = $c;
476 Because this operator produces an assignable result, using assignments
477 without parentheses will get you in trouble. For example, this:
479 $a % 2 ? $a += 10 : $a += 2
483 (($a % 2) ? ($a += 10) : $a) += 2
487 ($a % 2) ? ($a += 10) : ($a += 2)
489 That should probably be written more simply as:
491 $a += ($a % 2) ? 10 : 2;
493 =head2 Assignment Operators
495 "=" is the ordinary assignment operator.
497 Assignment operators work as in C. That is,
505 although without duplicating any side effects that dereferencing the lvalue
506 might trigger, such as from tie(). Other assignment operators work similarly.
507 The following are recognized:
514 Although these are grouped by family, they all have the precedence
517 Unlike in C, the assignment operator produces a valid lvalue. Modifying
518 an assignment is equivalent to doing the assignment and then modifying
519 the variable that was assigned to. This is useful for modifying
520 a copy of something, like this:
522 ($tmp = $global) =~ tr [A-Z] [a-z];
533 =head2 Comma Operator
535 Binary "," is the comma operator. In scalar context it evaluates
536 its left argument, throws that value away, then evaluates its right
537 argument and returns that value. This is just like C's comma operator.
539 In list context, it's just the list argument separator, and inserts
540 both its arguments into the list.
542 The =E<gt> digraph is mostly just a synonym for the comma operator. It's useful for
543 documenting arguments that come in pairs. As of release 5.001, it also forces
544 any word to the left of it to be interpreted as a string.
546 =head2 List Operators (Rightward)
548 On the right side of a list operator, it has very low precedence,
549 such that it controls all comma-separated expressions found there.
550 The only operators with lower precedence are the logical operators
551 "and", "or", and "not", which may be used to evaluate calls to list
552 operators without the need for extra parentheses:
554 open HANDLE, "filename"
555 or die "Can't open: $!\n";
557 See also discussion of list operators in L<Terms and List Operators (Leftward)>.
561 Unary "not" returns the logical negation of the expression to its right.
562 It's the equivalent of "!" except for the very low precedence.
566 Binary "and" returns the logical conjunction of the two surrounding
567 expressions. It's equivalent to && except for the very low
568 precedence. This means that it short-circuits: i.e., the right
569 expression is evaluated only if the left expression is true.
571 =head2 Logical or and Exclusive Or
573 Binary "or" returns the logical disjunction of the two surrounding
574 expressions. It's equivalent to || except for the very low precedence.
575 This makes it useful for control flow
577 print FH $data or die "Can't write to FH: $!";
579 This means that it short-circuits: i.e., the right expression is evaluated
580 only if the left expression is false. Due to its precedence, you should
581 probably avoid using this for assignment, only for control flow.
583 $a = $b or $c; # bug: this is wrong
584 ($a = $b) or $c; # really means this
585 $a = $b || $c; # better written this way
587 However, when it's a list-context assignment and you're trying to use
588 "||" for control flow, you probably need "or" so that the assignment
589 takes higher precedence.
591 @info = stat($file) || die; # oops, scalar sense of stat!
592 @info = stat($file) or die; # better, now @info gets its due
594 Then again, you could always use parentheses.
596 Binary "xor" returns the exclusive-OR of the two surrounding expressions.
597 It cannot short circuit, of course.
599 =head2 C Operators Missing From Perl
601 Here is what C has that Perl doesn't:
607 Address-of operator. (But see the "\" operator for taking a reference.)
611 Dereference-address operator. (Perl's prefix dereferencing
612 operators are typed: $, @, %, and &.)
616 Type-casting operator.
620 =head2 Quote and Quote-like Operators
622 While we usually think of quotes as literal values, in Perl they
623 function as operators, providing various kinds of interpolating and
624 pattern matching capabilities. Perl provides customary quote characters
625 for these behaviors, but also provides a way for you to choose your
626 quote character for any of them. In the following table, a C<{}> represents
627 any pair of delimiters you choose.
629 Customary Generic Meaning Interpolates
632 `` qx{} Command yes (unless '' is delimiter)
634 // m{} Pattern match yes (unless '' is delimiter)
635 qr{} Pattern yes (unless '' is delimiter)
636 s{}{} Substitution yes (unless '' is delimiter)
637 tr{}{} Transliteration no (but see below)
639 Non-bracketing delimiters use the same character fore and aft, but the four
640 sorts of brackets (round, angle, square, curly) will all nest, which means
649 Note, however, that this does not always work for quoting Perl code:
651 $s = q{ if($a eq "}") ... }; # WRONG
653 is a syntax error. The C<Text::Balanced> module on CPAN is able to do this
656 There can be whitespace between the operator and the quoting
657 characters, except when C<#> is being used as the quoting character.
658 C<q#foo#> is parsed as the string C<foo>, while C<q #foo#> is the
659 operator C<q> followed by a comment. Its argument will be taken
660 from the next line. This allows you to write:
662 s {foo} # Replace foo
665 For constructs that do interpolate, variables beginning with "C<$>"
666 or "C<@>" are interpolated, as are the following escape sequences. Within
667 a transliteration, the first eleven of these sequences may be used.
674 \a alarm (bell) (BEL)
676 \033 octal char (ESC)
678 \x{263a} wide hex char (SMILEY)
679 \c[ control char (ESC)
682 \l lowercase next char
683 \u uppercase next char
686 \E end case modification
687 \Q quote non-word characters till \E
689 If C<use locale> is in effect, the case map used by C<\l>, C<\L>, C<\u>
690 and C<\U> is taken from the current locale. See L<perllocale>. For
691 documentation of C<\N{name}>, see L<charnames>.
693 All systems use the virtual C<"\n"> to represent a line terminator,
694 called a "newline". There is no such thing as an unvarying, physical
695 newline character. It is only an illusion that the operating system,
696 device drivers, C libraries, and Perl all conspire to preserve. Not all
697 systems read C<"\r"> as ASCII CR and C<"\n"> as ASCII LF. For example,
698 on a Mac, these are reversed, and on systems without line terminator,
699 printing C<"\n"> may emit no actual data. In general, use C<"\n"> when
700 you mean a "newline" for your system, but use the literal ASCII when you
701 need an exact character. For example, most networking protocols expect
702 and prefer a CR+LF (C<"\012\015"> or C<"\cJ\cM">) for line terminators,
703 and although they often accept just C<"\012">, they seldom tolerate just
704 C<"\015">. If you get in the habit of using C<"\n"> for networking,
705 you may be burned some day.
707 You cannot include a literal C<$> or C<@> within a C<\Q> sequence.
708 An unescaped C<$> or C<@> interpolates the corresponding variable,
709 while escaping will cause the literal string C<\$> to be inserted.
710 You'll need to write something like C<m/\Quser\E\@\Qhost/>.
712 Patterns are subject to an additional level of interpretation as a
713 regular expression. This is done as a second pass, after variables are
714 interpolated, so that regular expressions may be incorporated into the
715 pattern from the variables. If this is not what you want, use C<\Q> to
716 interpolate a variable literally.
718 Apart from the behavior described above, Perl does not expand
719 multiple levels of interpolation. In particular, contrary to the
720 expectations of shell programmers, back-quotes do I<NOT> interpolate
721 within double quotes, nor do single quotes impede evaluation of
722 variables when used within double quotes.
724 =head2 Regexp Quote-Like Operators
726 Here are the quote-like operators that apply to pattern
727 matching and related activities.
733 This is just like the C</pattern/> search, except that it matches only
734 once between calls to the reset() operator. This is a useful
735 optimization when you want to see only the first occurrence of
736 something in each file of a set of files, for instance. Only C<??>
737 patterns local to the current package are reset.
741 # blank line between header and body
744 reset if eof; # clear ?? status for next file
747 This usage is vaguely depreciated, which means it just might possibly
748 be removed in some distant future version of Perl, perhaps somewhere
749 around the year 2168.
751 =item m/PATTERN/cgimosx
753 =item /PATTERN/cgimosx
755 Searches a string for a pattern match, and in scalar context returns
756 true if it succeeds, false if it fails. If no string is specified
757 via the C<=~> or C<!~> operator, the $_ string is searched. (The
758 string specified with C<=~> need not be an lvalue--it may be the
759 result of an expression evaluation, but remember the C<=~> binds
760 rather tightly.) See also L<perlre>. See L<perllocale> for
761 discussion of additional considerations that apply when C<use locale>
766 c Do not reset search position on a failed match when /g is in effect.
767 g Match globally, i.e., find all occurrences.
768 i Do case-insensitive pattern matching.
769 m Treat string as multiple lines.
770 o Compile pattern only once.
771 s Treat string as single line.
772 x Use extended regular expressions.
774 If "/" is the delimiter then the initial C<m> is optional. With the C<m>
775 you can use any pair of non-alphanumeric, non-whitespace characters
776 as delimiters. This is particularly useful for matching path names
777 that contain "/", to avoid LTS (leaning toothpick syndrome). If "?" is
778 the delimiter, then the match-only-once rule of C<?PATTERN?> applies.
779 If "'" is the delimiter, no interpolation is performed on the PATTERN.
781 PATTERN may contain variables, which will be interpolated (and the
782 pattern recompiled) every time the pattern search is evaluated, except
783 for when the delimiter is a single quote. (Note that C<$)> and C<$|>
784 might not be interpolated because they look like end-of-string tests.)
785 If you want such a pattern to be compiled only once, add a C</o> after
786 the trailing delimiter. This avoids expensive run-time recompilations,
787 and is useful when the value you are interpolating won't change over
788 the life of the script. However, mentioning C</o> constitutes a promise
789 that you won't change the variables in the pattern. If you change them,
790 Perl won't even notice. See also L<qr//>.
792 If the PATTERN evaluates to the empty string, the last
793 I<successfully> matched regular expression is used instead.
795 If the C</g> option is not used, C<m//> in list context returns a
796 list consisting of the subexpressions matched by the parentheses in the
797 pattern, i.e., (C<$1>, C<$2>, C<$3>...). (Note that here C<$1> etc. are
798 also set, and that this differs from Perl 4's behavior.) When there are
799 no parentheses in the pattern, the return value is the list C<(1)> for
800 success. With or without parentheses, an empty list is returned upon
805 open(TTY, '/dev/tty');
806 <TTY> =~ /^y/i && foo(); # do foo if desired
808 if (/Version: *([0-9.]*)/) { $version = $1; }
810 next if m#^/usr/spool/uucp#;
815 print if /$arg/o; # compile only once
818 if (($F1, $F2, $Etc) = ($foo =~ /^(\S+)\s+(\S+)\s*(.*)/))
820 This last example splits $foo into the first two words and the
821 remainder of the line, and assigns those three fields to $F1, $F2, and
822 $Etc. The conditional is true if any variables were assigned, i.e., if
825 The C</g> modifier specifies global pattern matching--that is,
826 matching as many times as possible within the string. How it behaves
827 depends on the context. In list context, it returns a list of the
828 substrings matched by any capturing parentheses in the regular
829 expression. If there are no parentheses, it returns a list of all
830 the matched strings, as if there were parentheses around the whole
833 In scalar context, each execution of C<m//g> finds the next match,
834 returning true if it matches, and false if there is no further match.
835 The position after the last match can be read or set using the pos()
836 function; see L<perlfunc/pos>. A failed match normally resets the
837 search position to the beginning of the string, but you can avoid that
838 by adding the C</c> modifier (e.g. C<m//gc>). Modifying the target
839 string also resets the search position.
841 You can intermix C<m//g> matches with C<m/\G.../g>, where C<\G> is a
842 zero-width assertion that matches the exact position where the previous
843 C<m//g>, if any, left off. The C<\G> assertion is not supported without
844 the C</g> modifier. (Currently, without C</g>, C<\G> behaves just like
845 C<\A>, but that's accidental and may change in the future.)
850 ($one,$five,$fifteen) = (`uptime` =~ /(\d+\.\d+)/g);
853 $/ = ""; $* = 1; # $* deprecated in modern perls
854 while (defined($paragraph = <>)) {
855 while ($paragraph =~ /[a-z]['")]*[.!?]+['")]*\s/g) {
859 print "$sentences\n";
861 # using m//gc with \G
865 print $1 while /(o)/gc; print "', pos=", pos, "\n";
867 print $1 if /\G(q)/gc; print "', pos=", pos, "\n";
869 print $1 while /(p)/gc; print "', pos=", pos, "\n";
872 The last example should print:
881 A useful idiom for C<lex>-like scanners is C</\G.../gc>. You can
882 combine several regexps like this to process a string part-by-part,
883 doing different actions depending on which regexp matched. Each
884 regexp tries to match where the previous one leaves off.
887 $url = new URI::URL "http://www/"; die if $url eq "xXx";
891 print(" digits"), redo LOOP if /\G\d+\b[,.;]?\s*/gc;
892 print(" lowercase"), redo LOOP if /\G[a-z]+\b[,.;]?\s*/gc;
893 print(" UPPERCASE"), redo LOOP if /\G[A-Z]+\b[,.;]?\s*/gc;
894 print(" Capitalized"), redo LOOP if /\G[A-Z][a-z]+\b[,.;]?\s*/gc;
895 print(" MiXeD"), redo LOOP if /\G[A-Za-z]+\b[,.;]?\s*/gc;
896 print(" alphanumeric"), redo LOOP if /\G[A-Za-z0-9]+\b[,.;]?\s*/gc;
897 print(" line-noise"), redo LOOP if /\G[^A-Za-z0-9]+/gc;
898 print ". That's all!\n";
901 Here is the output (split into several lines):
903 line-noise lowercase line-noise lowercase UPPERCASE line-noise
904 UPPERCASE line-noise lowercase line-noise lowercase line-noise
905 lowercase lowercase line-noise lowercase lowercase line-noise
906 MiXeD line-noise. That's all!
912 A single-quoted, literal string. A backslash represents a backslash
913 unless followed by the delimiter or another backslash, in which case
914 the delimiter or backslash is interpolated.
916 $foo = q!I said, "You said, 'She said it.'"!;
917 $bar = q('This is it.');
918 $baz = '\n'; # a two-character string
924 A double-quoted, interpolated string.
927 (*** The previous line contains the naughty word "$1".\n)
928 if /\b(tcl|java|python)\b/i; # :-)
929 $baz = "\n"; # a one-character string
931 =item qr/STRING/imosx
933 This operators quotes--and compiles--its I<STRING> as a regular
934 expression. I<STRING> is interpolated the same way as I<PATTERN>
935 in C<m/PATTERN/>. If "'" is used as the delimiter, no interpolation
936 is done. Returns a Perl value which may be used instead of the
937 corresponding C</STRING/imosx> expression.
941 $rex = qr/my.STRING/is;
948 The result may be used as a subpattern in a match:
951 $string =~ /foo${re}bar/; # can be interpolated in other patterns
952 $string =~ $re; # or used standalone
953 $string =~ /$re/; # or this way
955 Since Perl may compile the pattern at the moment of execution of qr()
956 operator, using qr() may have speed advantages in some situations,
957 notably if the result of qr() is used standalone:
960 my $patterns = shift;
961 my @compiled = map qr/$_/i, @$patterns;
964 foreach my $pat @compiled {
965 $success = 1, last if /$pat/;
971 Precompilation of the pattern into an internal representation at
972 the moment of qr() avoids a need to recompile the pattern every
973 time a match C</$pat/> is attempted. (Perl has many other internal
974 optimizations, but none would be triggered in the above example if
975 we did not use qr() operator.)
979 i Do case-insensitive pattern matching.
980 m Treat string as multiple lines.
981 o Compile pattern only once.
982 s Treat string as single line.
983 x Use extended regular expressions.
985 See L<perlre> for additional information on valid syntax for STRING, and
986 for a detailed look at the semantics of regular expressions.
992 A string which is (possibly) interpolated and then executed as a system
993 command with C</bin/sh> or its equivalent. Shell wildcards, pipes,
994 and redirections will be honored. The collected standard output of the
995 command is returned; standard error is unaffected. In scalar context,
996 it comes back as a single (potentially multi-line) string. In list
997 context, returns a list of lines (however you've defined lines with $/
998 or $INPUT_RECORD_SEPARATOR).
1000 Because backticks do not affect standard error, use shell file descriptor
1001 syntax (assuming the shell supports this) if you care to address this.
1002 To capture a command's STDERR and STDOUT together:
1004 $output = `cmd 2>&1`;
1006 To capture a command's STDOUT but discard its STDERR:
1008 $output = `cmd 2>/dev/null`;
1010 To capture a command's STDERR but discard its STDOUT (ordering is
1013 $output = `cmd 2>&1 1>/dev/null`;
1015 To exchange a command's STDOUT and STDERR in order to capture the STDERR
1016 but leave its STDOUT to come out the old STDERR:
1018 $output = `cmd 3>&1 1>&2 2>&3 3>&-`;
1020 To read both a command's STDOUT and its STDERR separately, it's easiest
1021 and safest to redirect them separately to files, and then read from those
1022 files when the program is done:
1024 system("program args 1>/tmp/program.stdout 2>/tmp/program.stderr");
1026 Using single-quote as a delimiter protects the command from Perl's
1027 double-quote interpolation, passing it on to the shell instead:
1029 $perl_info = qx(ps $$); # that's Perl's $$
1030 $shell_info = qx'ps $$'; # that's the new shell's $$
1032 How that string gets evaluated is entirely subject to the command
1033 interpreter on your system. On most platforms, you will have to protect
1034 shell metacharacters if you want them treated literally. This is in
1035 practice difficult to do, as it's unclear how to escape which characters.
1036 See L<perlsec> for a clean and safe example of a manual fork() and exec()
1037 to emulate backticks safely.
1039 On some platforms (notably DOS-like ones), the shell may not be
1040 capable of dealing with multiline commands, so putting newlines in
1041 the string may not get you what you want. You may be able to evaluate
1042 multiple commands in a single line by separating them with the command
1043 separator character, if your shell supports that (e.g. C<;> on many Unix
1044 shells; C<&> on the Windows NT C<cmd> shell).
1046 Beware that some command shells may place restrictions on the length
1047 of the command line. You must ensure your strings don't exceed this
1048 limit after any necessary interpolations. See the platform-specific
1049 release notes for more details about your particular environment.
1051 Using this operator can lead to programs that are difficult to port,
1052 because the shell commands called vary between systems, and may in
1053 fact not be present at all. As one example, the C<type> command under
1054 the POSIX shell is very different from the C<type> command under DOS.
1055 That doesn't mean you should go out of your way to avoid backticks
1056 when they're the right way to get something done. Perl was made to be
1057 a glue language, and one of the things it glues together is commands.
1058 Just understand what you're getting yourself into.
1060 See L<"I/O Operators"> for more discussion.
1064 Evaluates to a list of the words extracted out of STRING, using embedded
1065 whitespace as the word delimiters. It can be understood as being roughly
1068 split(' ', q/STRING/);
1070 the difference being that it generates a real list at compile time. So
1075 is exactly equivalent to the list:
1077 ('foo', 'bar', 'baz')
1079 Some frequently seen examples:
1081 use POSIX qw( setlocale localeconv )
1082 @EXPORT = qw( foo bar baz );
1084 A common mistake is to try to separate the words with comma or to
1085 put comments into a multi-line C<qw>-string. For this reason, the
1086 B<-w> switch (that is, the C<$^W> variable) produces warnings if
1087 the STRING contains the "," or the "#" character.
1089 =item s/PATTERN/REPLACEMENT/egimosx
1091 Searches a string for a pattern, and if found, replaces that pattern
1092 with the replacement text and returns the number of substitutions
1093 made. Otherwise it returns false (specifically, the empty string).
1095 If no string is specified via the C<=~> or C<!~> operator, the C<$_>
1096 variable is searched and modified. (The string specified with C<=~> must
1097 be scalar variable, an array element, a hash element, or an assignment
1098 to one of those, i.e., an lvalue.)
1100 If the delimiter chosen is a single quote, no interpolation is
1101 done on either the PATTERN or the REPLACEMENT. Otherwise, if the
1102 PATTERN contains a $ that looks like a variable rather than an
1103 end-of-string test, the variable will be interpolated into the pattern
1104 at run-time. If you want the pattern compiled only once the first time
1105 the variable is interpolated, use the C</o> option. If the pattern
1106 evaluates to the empty string, the last successfully executed regular
1107 expression is used instead. See L<perlre> for further explanation on these.
1108 See L<perllocale> for discussion of additional considerations that apply
1109 when C<use locale> is in effect.
1113 e Evaluate the right side as an expression.
1114 g Replace globally, i.e., all occurrences.
1115 i Do case-insensitive pattern matching.
1116 m Treat string as multiple lines.
1117 o Compile pattern only once.
1118 s Treat string as single line.
1119 x Use extended regular expressions.
1121 Any non-alphanumeric, non-whitespace delimiter may replace the
1122 slashes. If single quotes are used, no interpretation is done on the
1123 replacement string (the C</e> modifier overrides this, however). Unlike
1124 Perl 4, Perl 5 treats backticks as normal delimiters; the replacement
1125 text is not evaluated as a command. If the
1126 PATTERN is delimited by bracketing quotes, the REPLACEMENT has its own
1127 pair of quotes, which may or may not be bracketing quotes, e.g.,
1128 C<s(foo)(bar)> or C<sE<lt>fooE<gt>/bar/>. A C</e> will cause the
1129 replacement portion to be interpreted as a full-fledged Perl expression
1130 and eval()ed right then and there. It is, however, syntax checked at
1135 s/\bgreen\b/mauve/g; # don't change wintergreen
1137 $path =~ s|/usr/bin|/usr/local/bin|;
1139 s/Login: $foo/Login: $bar/; # run-time pattern
1141 ($foo = $bar) =~ s/this/that/; # copy first, then change
1143 $count = ($paragraph =~ s/Mister\b/Mr./g); # get change-count
1146 s/\d+/$&*2/e; # yields 'abc246xyz'
1147 s/\d+/sprintf("%5d",$&)/e; # yields 'abc 246xyz'
1148 s/\w/$& x 2/eg; # yields 'aabbcc 224466xxyyzz'
1150 s/%(.)/$percent{$1}/g; # change percent escapes; no /e
1151 s/%(.)/$percent{$1} || $&/ge; # expr now, so /e
1152 s/^=(\w+)/&pod($1)/ge; # use function call
1154 # expand variables in $_, but dynamics only, using
1155 # symbolic dereferencing
1158 # /e's can even nest; this will expand
1159 # any embedded scalar variable (including lexicals) in $_
1162 # Delete (most) C comments.
1164 /\* # Match the opening delimiter.
1165 .*? # Match a minimal number of characters.
1166 \*/ # Match the closing delimiter.
1169 s/^\s*(.*?)\s*$/$1/; # trim white space in $_, expensively
1171 for ($variable) { # trim white space in $variable, cheap
1176 s/([^ ]*) *([^ ]*)/$2 $1/; # reverse 1st two fields
1178 Note the use of $ instead of \ in the last example. Unlike
1179 B<sed>, we use the \E<lt>I<digit>E<gt> form in only the left hand side.
1180 Anywhere else it's $E<lt>I<digit>E<gt>.
1182 Occasionally, you can't use just a C</g> to get all the changes
1183 to occur that you might want. Here are two common cases:
1185 # put commas in the right places in an integer
1186 1 while s/(\d)(\d\d\d)(?!\d)/$1,$2/g;
1188 # expand tabs to 8-column spacing
1189 1 while s/\t+/' ' x (length($&)*8 - length($`)%8)/e;
1191 =item tr/SEARCHLIST/REPLACEMENTLIST/cdsUC
1193 =item y/SEARCHLIST/REPLACEMENTLIST/cdsUC
1195 Transliterates all occurrences of the characters found in the search list
1196 with the corresponding character in the replacement list. It returns
1197 the number of characters replaced or deleted. If no string is
1198 specified via the =~ or !~ operator, the $_ string is transliterated. (The
1199 string specified with =~ must be a scalar variable, an array element, a
1200 hash element, or an assignment to one of those, i.e., an lvalue.)
1202 A character range may be specified with a hyphen, so C<tr/A-J/0-9/>
1203 does the same replacement as C<tr/ACEGIBDFHJ/0246813579/>.
1204 For B<sed> devotees, C<y> is provided as a synonym for C<tr>. If the
1205 SEARCHLIST is delimited by bracketing quotes, the REPLACEMENTLIST has
1206 its own pair of quotes, which may or may not be bracketing quotes,
1207 e.g., C<tr[A-Z][a-z]> or C<tr(+\-*/)/ABCD/>.
1209 Note also that the whole range idea is rather unportable between
1210 character sets--and even within character sets they may cause results
1211 you probably didn't expect. A sound principle is to use only ranges
1212 that begin from and end at either alphabets of equal case (a-e, A-E),
1213 or digits (0-4). Anything else is unsafe. If in doubt, spell out the
1214 character sets in full.
1218 c Complement the SEARCHLIST.
1219 d Delete found but unreplaced characters.
1220 s Squash duplicate replaced characters.
1221 U Translate to/from UTF-8.
1222 C Translate to/from 8-bit char (octet).
1224 If the C</c> modifier is specified, the SEARCHLIST character set
1225 is complemented. If the C</d> modifier is specified, any characters
1226 specified by SEARCHLIST not found in REPLACEMENTLIST are deleted.
1227 (Note that this is slightly more flexible than the behavior of some
1228 B<tr> programs, which delete anything they find in the SEARCHLIST,
1229 period.) If the C</s> modifier is specified, sequences of characters
1230 that were transliterated to the same character are squashed down
1231 to a single instance of the character.
1233 If the C</d> modifier is used, the REPLACEMENTLIST is always interpreted
1234 exactly as specified. Otherwise, if the REPLACEMENTLIST is shorter
1235 than the SEARCHLIST, the final character is replicated till it is long
1236 enough. If the REPLACEMENTLIST is empty, the SEARCHLIST is replicated.
1237 This latter is useful for counting characters in a class or for
1238 squashing character sequences in a class.
1240 The first C</U> or C</C> modifier applies to the left side of the translation.
1241 The second one applies to the right side. If present, these modifiers override
1242 the current utf8 state.
1246 $ARGV[1] =~ tr/A-Z/a-z/; # canonicalize to lower case
1248 $cnt = tr/*/*/; # count the stars in $_
1250 $cnt = $sky =~ tr/*/*/; # count the stars in $sky
1252 $cnt = tr/0-9//; # count the digits in $_
1254 tr/a-zA-Z//s; # bookkeeper -> bokeper
1256 ($HOST = $host) =~ tr/a-z/A-Z/;
1258 tr/a-zA-Z/ /cs; # change non-alphas to single space
1261 [\000-\177]; # delete 8th bit
1263 tr/\0-\xFF//CU; # change Latin-1 to Unicode
1264 tr/\0-\x{FF}//UC; # change Unicode to Latin-1
1266 If multiple transliterations are given for a character, only the
1271 will transliterate any A to X.
1273 Because the transliteration table is built at compile time, neither
1274 the SEARCHLIST nor the REPLACEMENTLIST are subjected to double quote
1275 interpolation. That means that if you want to use variables, you
1278 eval "tr/$oldlist/$newlist/";
1281 eval "tr/$oldlist/$newlist/, 1" or die $@;
1285 =head2 Gory details of parsing quoted constructs
1287 When presented with something that might have several different
1288 interpretations, Perl uses the B<DWIM> (that's "Do What I Mean")
1289 principle to pick the most probable interpretation. This strategy
1290 is so successful that Perl programmers often do not suspect the
1291 ambivalence of what they write. But from time to time, Perl's
1292 notions differ substantially from what the author honestly meant.
1294 This section hopes to clarify how Perl handles quoted constructs.
1295 Although the most common reason to learn this is to unravel labyrinthine
1296 regular expressions, because the initial steps of parsing are the
1297 same for all quoting operators, they are all discussed together.
1299 The most important Perl parsing rule is the first one discussed
1300 below: when processing a quoted construct, Perl first finds the end
1301 of that construct, then interprets its contents. If you understand
1302 this rule, you may skip the rest of this section on the first
1303 reading. The other rules are likely to contradict the user's
1304 expectations much less frequently than this first one.
1306 Some passes discussed below are performed concurrently, but because
1307 their results are the same, we consider them individually. For different
1308 quoting constructs, Perl performs different numbers of passes, from
1309 one to five, but these passes are always performed in the same order.
1313 =item Finding the end
1315 The first pass is finding the end of the quoted construct, whether
1316 it be a multicharacter delimiter C<"\nEOF\n"> in the C<<<EOF>
1317 construct, a C</> that terminates a C<qq//> construct, a C<]> which
1318 terminates C<qq[]> construct, or a C<E<gt>> which terminates a
1319 fileglob started with C<E<lt>>.
1321 When searching for single-character non-pairing delimiters, such
1322 as C</>, combinations of C<\\> and C<\/> are skipped. However,
1323 when searching for single-character pairing delimiter like C<[>,
1324 combinations of C<\\>, C<\]>, and C<\[> are all skipped, and nested
1325 C<[>, C<]> are skipped as well. When searching for multicharacter
1326 delimiters, nothing is skipped.
1328 For constructs with three-part delimiters (C<s///>, C<y///>, and
1329 C<tr///>), the search is repeated once more.
1331 During this search no attention is paid to the semantics of the construct.
1334 "$hash{"$foo/$bar"}"
1339 bar # NOT a comment, this slash / terminated m//!
1342 do not form legal quoted expressions. The quoted part ends on the
1343 first C<"> and C</>, and the rest happens to be a syntax error.
1344 Because the slash that terminated C<m//> was followed by a C<SPACE>,
1345 the example above is not C<m//x>, but rather C<m//> with no C</x>
1346 modifier. So the embedded C<#> is interpreted as a literal C<#>.
1348 =item Removal of backslashes before delimiters
1350 During the second pass, text between the starting and ending
1351 delimiters is copied to a safe location, and the C<\> is removed
1352 from combinations consisting of C<\> and delimiter--or delimiters,
1353 meaning both starting and ending delimiters will should these differ.
1354 This removal does not happen for multi-character delimiters.
1355 Note that the combination C<\\> is left intact, just as it was.
1357 Starting from this step no information about the delimiters is
1362 The next step is interpolation in the text obtained, which is now
1363 delimiter-independent. There are four different cases.
1367 =item C<<<'EOF'>, C<m''>, C<s'''>, C<tr///>, C<y///>
1369 No interpolation is performed.
1373 The only interpolation is removal of C<\> from pairs C<\\>.
1375 =item C<"">, C<``>, C<qq//>, C<qx//>, C<<file*globE<gt>>
1377 C<\Q>, C<\U>, C<\u>, C<\L>, C<\l> (possibly paired with C<\E>) are
1378 converted to corresponding Perl constructs. Thus, C<"$foo\Qbaz$bar">
1379 is converted to C<$foo . (quotemeta("baz" . $bar))> internally.
1380 The other combinations are replaced with appropriate expansions.
1382 Let it be stressed that I<whatever falls between C<\Q> and C<\E>>
1383 is interpolated in the usual way. Something like C<"\Q\\E"> has
1384 no C<\E> inside. instead, it has C<\Q>, C<\\>, and C<E>, so the
1385 result is the same as for C<"\\\\E">. As a general rule, backslashes
1386 between C<\Q> and C<\E> may lead to counterintuitive results. So,
1387 C<"\Q\t\E"> is converted to C<quotemeta("\t")>, which is the same
1388 as C<"\\\t"> (since TAB is not alphanumeric). Note also that:
1393 may be closer to the conjectural I<intention> of the writer of C<"\Q\t\E">.
1395 Interpolated scalars and arrays are converted internally to the C<join> and
1396 C<.> catentation operations. Thus, C<"$foo XXX '@arr'"> becomes:
1398 $foo . " XXX '" . (join $", @arr) . "'";
1400 All operations above are performed simultaneously, left to right.
1402 Because the result of C<"\Q STRING \E"> has all metacharacters
1403 quoted, there is no way to insert a literal C<$> or C<@> inside a
1404 C<\Q\E> pair. If protected by C<\>, C<$> will be quoted to became
1405 C<"\\\$">; if not, it is interpreted as the start of an interpolated
1408 Note also that the interpolation code needs to make a decision on
1409 where the interpolated scalar ends. For instance, whether
1410 C<"a $b -E<gt> {c}"> really means:
1412 "a " . $b . " -> {c}";
1418 Most of the time, the longest possible text that does not include
1419 spaces between components and which contains matching braces or
1420 brackets. because the outcome may be determined by voting based
1421 on heuristic estimators, the result is not strictly predictable.
1422 Fortunately, it's usually correct for ambiguous cases.
1424 =item C<?RE?>, C</RE/>, C<m/RE/>, C<s/RE/foo/>,
1426 Processing of C<\Q>, C<\U>, C<\u>, C<\L>, C<\l>, and interpolation
1427 happens (almost) as with C<qq//> constructs, but the substitution
1428 of C<\> followed by RE-special chars (including C<\>) is not
1429 performed. Moreover, inside C<(?{BLOCK})>, C<(?# comment )>, and
1430 a C<#>-comment in a C<//x>-regular expression, no processing is
1431 performed whatsoever. This is the first step at which the presence
1432 of the C<//x> modifier is relevant.
1434 Interpolation has several quirks: C<$|>, C<$(>, and C<$)> are not
1435 interpolated, and constructs C<$var[SOMETHING]> are voted (by several
1436 different estimators) to be either an array element or C<$var>
1437 followed by an RE alternative. This is where the notation
1438 C<${arr[$bar]}> comes handy: C</${arr[0-9]}/> is interpreted as
1439 array element C<-9>, not as a regular expression from the variable
1440 C<$arr> followed by a digit, which would be the interpretation of
1441 C</$arr[0-9]/>. Since voting among different estimators may occur,
1442 the result is not predictable.
1444 It is at this step that C<\1> is begrudgingly converted to C<$1> in
1445 the replacement text of C<s///> to correct the incorrigible
1446 I<sed> hackers who haven't picked up the saner idiom yet. A warning
1447 is emitted if the B<-w> command-line flag (that is, the C<$^W> variable)
1450 The lack of processing of C<\\> creates specific restrictions on
1451 the post-processed text. If the delimiter is C</>, one cannot get
1452 the combination C<\/> into the result of this step. C</> will
1453 finish the regular expression, C<\/> will be stripped to C</> on
1454 the previous step, and C<\\/> will be left as is. Because C</> is
1455 equivalent to C<\/> inside a regular expression, this does not
1456 matter unless the delimiter happens to be character special to the
1457 RE engine, such as in C<s*foo*bar*>, C<m[foo]>, or C<?foo?>; or an
1458 alphanumeric char, as in:
1462 In the RE above, which is intentionally obfuscated for illustration, the
1463 delimiter is C<m>, the modifier is C<mx>, and after backslash-removal the
1464 RE is the same as for C<m/ ^ a s* b /mx>). There's more than one
1465 reason you're encouraged to restrict your delimiters to non-alphanumeric,
1466 non-whitespace choices.
1470 This step is the last one for all constructs except regular expressions,
1471 which are processed further.
1473 =item Interpolation of regular expressions
1475 Previous steps were performed during the compilation of Perl code,
1476 but this one happens at run time--although it may be optimized to
1477 be calculated at compile time if appropriate. After preprocessing
1478 described above, and possibly after evaluation if catenation,
1479 joining, casing translation, or metaquoting are involved, the
1480 resulting I<string> is passed to the RE engine for compilation.
1482 Whatever happens in the RE engine might be better discussed in L<perlre>,
1483 but for the sake of continuity, we shall do so here.
1485 This is another step where the presence of the C<//x> modifier is
1486 relevant. The RE engine scans the string from left to right and
1487 converts it to a finite automaton.
1489 Backslashed characters are either replaced with corresponding
1490 literal strings (as with C<\{>), or else they generate special nodes
1491 in the finite automaton (as with C<\b>). Characters special to the
1492 RE engine (such as C<|>) generate corresponding nodes or groups of
1493 nodes. C<(?#...)> comments are ignored. All the rest is either
1494 converted to literal strings to match, or else is ignored (as is
1495 whitespace and C<#>-style comments if C<//x> is present).
1497 Parsing of the bracketed character class construct, C<[...]>, is
1498 rather different than the rule used for the rest of the pattern.
1499 The terminator of this construct is found using the same rules as
1500 for finding the terminator of a C<{}>-delimited construct, the only
1501 exception being that C<]> immediately following C<[> is treated as
1502 though preceded by a backslash. Similarly, the terminator of
1503 C<(?{...})> is found using the same rules as for finding the
1504 terminator of a C<{}>-delimited construct.
1506 It is possible to inspect both the string given to RE engine and the
1507 resulting finite automaton. See the arguments C<debug>/C<debugcolor>
1508 in the C<use L<re>> pragma, as well as Perl's B<-Dr> command-line
1509 switch documented in L<perlrun/Switches>.
1511 =item Optimization of regular expressions
1513 This step is listed for completeness only. Since it does not change
1514 semantics, details of this step are not documented and are subject
1515 to change without notice. This step is performed over the finite
1516 automaton that was generated during the previous pass.
1518 It is at this stage that C<split()> silently optimizes C</^/> to
1523 =head2 I/O Operators
1525 There are several I/O operators you should know about.
1527 A string enclosed by backticks (grave accents) first undergoes
1528 double-quote interpolation. It is then interpreted as an external
1529 command, and the output of that command is the value of the
1531 string consisting of all output is returned. In list context, a
1532 list of values is returned, one per line of output. (You can set
1533 C<$/> to use a different line terminator.) The command is executed
1534 each time the pseudo-literal is evaluated. The status value of the
1535 command is returned in C<$?> (see L<perlvar> for the interpretation
1536 of C<$?>). Unlike in B<csh>, no translation is done on the return
1537 data--newlines remain newlines. Unlike in any of the shells, single
1538 quotes do not hide variable names in the command from interpretation.
1539 To pass a literal dollar-sign through to the shell you need to hide
1540 it with a backslash. The generalized form of backticks is C<qx//>.
1541 (Because backticks always undergo shell expansion as well, see
1542 L<perlsec> for security concerns.)
1544 In scalar context, evaluating a filehandle in angle brackets yields
1545 the next line from that file (the newline, if any, included), or
1546 C<undef> at end-of-file or on error. When C<$/> is set to C<undef>
1547 (sometimes known as file-slurp mode) and the file is empty, it
1548 returns C<''> the first time, followed by C<undef> subsequently.
1550 Ordinarily you must assign the returned value to a variable, but
1551 there is one situation where an automatic assignment happens. If
1552 and only if the input symbol is the only thing inside the conditional
1553 of a C<while> statement (even if disguised as a C<for(;;)> loop),
1554 the value is automatically assigned to the global variable $_,
1555 destroying whatever was there previously. (This may seem like an
1556 odd thing to you, but you'll use the construct in almost every Perl
1557 script you write.) The $_ variables is not implicitly localized.
1558 You'll have to put a C<local $_;> before the loop if you want that
1561 The following lines are equivalent:
1563 while (defined($_ = <STDIN>)) { print; }
1564 while ($_ = <STDIN>) { print; }
1565 while (<STDIN>) { print; }
1566 for (;<STDIN>;) { print; }
1567 print while defined($_ = <STDIN>);
1568 print while ($_ = <STDIN>);
1569 print while <STDIN>;
1571 This also behaves similarly, but avoids $_ :
1573 while (my $line = <STDIN>) { print $line }
1575 In these loop constructs, the assigned value (whether assignment
1576 is automatic or explicit) is then tested to see whether it is
1577 defined. The defined test avoids problems where line has a string
1578 value that would be treated as false by Perl, for example a "" or
1579 a "0" with no trailing newline. If you really mean for such values
1580 to terminate the loop, they should be tested for explicitly:
1582 while (($_ = <STDIN>) ne '0') { ... }
1583 while (<STDIN>) { last unless $_; ... }
1585 In other boolean contexts, C<E<lt>I<filehandle>E<gt>> without an
1586 explicit C<defined> test or comparison elicit a warning if the B<-w>
1587 command-line switch (the C<$^W> variable) is in effect.
1589 The filehandles STDIN, STDOUT, and STDERR are predefined. (The
1590 filehandles C<stdin>, C<stdout>, and C<stderr> will also work except
1591 in packages, where they would be interpreted as local identifiers
1592 rather than global.) Additional filehandles may be created with
1593 the open() function, amongst others. See L<perlopentut> and
1594 L<perlfunc/open> for details on this.
1596 If a E<lt>FILEHANDLEE<gt> is used in a context that is looking for
1597 a list, a list comprising all input lines is returned, one line per
1598 list element. It's easy to grow to a rather large data space this
1599 way, so use with care.
1601 E<lt>FILEHANDLEE<gt> may also be spelled C<readline(*FILEHANDLE)>.
1602 See L<perlfunc/readline>.
1604 The null filehandle E<lt>E<gt> is special: it can be used to emulate the
1605 behavior of B<sed> and B<awk>. Input from E<lt>E<gt> comes either from
1606 standard input, or from each file listed on the command line. Here's
1607 how it works: the first time E<lt>E<gt> is evaluated, the @ARGV array is
1608 checked, and if it is empty, C<$ARGV[0]> is set to "-", which when opened
1609 gives you standard input. The @ARGV array is then processed as a list
1610 of filenames. The loop
1613 ... # code for each line
1616 is equivalent to the following Perl-like pseudo code:
1618 unshift(@ARGV, '-') unless @ARGV;
1619 while ($ARGV = shift) {
1622 ... # code for each line
1626 except that it isn't so cumbersome to say, and will actually work.
1627 It really does shift the @ARGV array and put the current filename
1628 into the $ARGV variable. It also uses filehandle I<ARGV>
1629 internally--E<lt>E<gt> is just a synonym for E<lt>ARGVE<gt>, which
1630 is magical. (The pseudo code above doesn't work because it treats
1631 E<lt>ARGVE<gt> as non-magical.)
1633 You can modify @ARGV before the first E<lt>E<gt> as long as the array ends up
1634 containing the list of filenames you really want. Line numbers (C<$.>)
1635 continue as though the input were one big happy file. See the example
1636 in L<perlfunc/eof> for how to reset line numbers on each file.
1638 If you want to set @ARGV to your own list of files, go right ahead.
1639 This sets @ARGV to all plain text files if no @ARGV was given:
1641 @ARGV = grep { -f && -T } glob('*') unless @ARGV;
1643 You can even set them to pipe commands. For example, this automatically
1644 filters compressed arguments through B<gzip>:
1646 @ARGV = map { /\.(gz|Z)$/ ? "gzip -dc < $_ |" : $_ } @ARGV;
1648 If you want to pass switches into your script, you can use one of the
1649 Getopts modules or put a loop on the front like this:
1651 while ($_ = $ARGV[0], /^-/) {
1654 if (/^-D(.*)/) { $debug = $1 }
1655 if (/^-v/) { $verbose++ }
1656 # ... # other switches
1660 # ... # code for each line
1663 The E<lt>E<gt> symbol will return C<undef> for end-of-file only once.
1664 If you call it again after this, it will assume you are processing another
1665 @ARGV list, and if you haven't set @ARGV, will read input from STDIN.
1667 If angle brackets contain is a simple scalar variable (e.g.,
1668 E<lt>$fooE<gt>), then that variable contains the name of the
1669 filehandle to input from, or its typeglob, or a reference to the
1675 If what's within the angle brackets is neither a filehandle nor a simple
1676 scalar variable containing a filehandle name, typeglob, or typeglob
1677 reference, it is interpreted as a filename pattern to be globbed, and
1678 either a list of filenames or the next filename in the list is returned,
1679 depending on context. This distinction is determined on syntactic
1680 grounds alone. That means C<E<lt>$xE<gt>> is always a readline() from
1681 an indirect handle, but C<E<lt>$hash{key}E<gt>> is always a glob().
1682 That's because $x is a simple scalar variable, but C<$hash{key}> is
1683 not--it's a hash element.
1685 One level of double-quote interpretation is done first, but you can't
1686 say C<E<lt>$fooE<gt>> because that's an indirect filehandle as explained
1687 in the previous paragraph. (In older versions of Perl, programmers
1688 would insert curly brackets to force interpretation as a filename glob:
1689 C<E<lt>${foo}E<gt>>. These days, it's considered cleaner to call the
1690 internal function directly as C<glob($foo)>, which is probably the right
1691 way to have done it in the first place.) For example:
1699 open(FOO, "echo *.c | tr -s ' \t\r\f' '\\012\\012\\012\\012'|");
1705 In fact, it's currently implemented that way, but this is expected
1706 to be made completely internal in the near future. (Which means
1707 it will not work on filenames with spaces in them unless you have
1708 csh(1) on your machine.) Of course, the shortest way to do the
1713 Because globbing currently invokes a shell, it's often faster to
1714 call readdir() yourself and do your own grep() on the filenames.
1715 Furthermore, due to its current implementation of using a shell,
1716 the glob() routine may get "Arg list too long" errors (unless you've
1717 installed tcsh(1L) as F</bin/csh> or hacked your F<config.sh>).
1719 A (file)glob evaluates its (embedded) argument only when it is
1720 starting a new list. All values must be read before it will start
1721 over. In list context, this isn't important because you automatically
1722 get them all anyway. However, in scalar context the operator returns
1723 the next value each time it's called, or C
1724 run out. As with filehandle reads, an automatic C<defined> is
1725 generated when the glob occurs in the test part of a C<while>,
1726 because legal glob returns (e.g. a file called F<0>) would otherwise
1727 terminate the loop. Again, C<undef> is returned only once. So if
1728 you're expecting a single value from a glob, it is much better to
1731 ($file) = <blurch*>;
1737 because the latter will alternate between returning a filename and
1740 It you're trying to do variable interpolation, it's definitely better
1741 to use the glob() function, because the older notation can cause people
1742 to become confused with the indirect filehandle notation.
1744 @files = glob("$dir/*.[ch]");
1745 @files = glob($files[$i]);
1747 =head2 Constant Folding
1749 Like C, Perl does a certain amount of expression evaluation at
1750 compile time whenever it determines that all arguments to an
1751 operator are static and have no side effects. In particular, string
1752 concatenation happens at compile time between literals that don't do
1753 variable substitution. Backslash interpolation also happens at
1754 compile time. You can say
1756 'Now is the time for all' . "\n" .
1757 'good men to come to.'
1759 and this all reduces to one string internally. Likewise, if
1762 foreach $file (@filenames) {
1763 if (-s $file > 5 + 100 * 2**16) { }
1766 the compiler will precompute the number which that expression
1767 represents so that the interpreter won't have to.
1769 =head2 Bitwise String Operators
1771 Bitstrings of any size may be manipulated by the bitwise operators
1774 If the operands to a binary bitwise op are strings of different
1775 sizes, B<|> and B<^> ops act as though the shorter operand had
1776 additional zero bits on the right, while the B<&> op acts as though
1777 the longer operand were truncated to the length of the shorter.
1778 The granularity for such extension or truncation is one or more
1781 # ASCII-based examples
1782 print "j p \n" ^ " a h"; # prints "JAPH\n"
1783 print "JA" | " ph\n"; # prints "japh\n"
1784 print "japh\nJunk" & '_____'; # prints "JAPH\n";
1785 print 'p N$' ^ " E<H\n"; # prints "Perl\n";
1787 If you are intending to manipulate bitstrings, be certain that
1788 you're supplying bitstrings: If an operand is a number, that will imply
1789 a B<numeric> bitwise operation. You may explicitly show which type of
1790 operation you intend by using C<""> or C<0+>, as in the examples below.
1792 $foo = 150 | 105 ; # yields 255 (0x96 | 0x69 is 0xFF)
1793 $foo = '150' | 105 ; # yields 255
1794 $foo = 150 | '105'; # yields 255
1795 $foo = '150' | '105'; # yields string '155' (under ASCII)
1797 $baz = 0+$foo & 0+$bar; # both ops explicitly numeric
1798 $biz = "$foo" ^ "$bar"; # both ops explicitly stringy
1800 See L<perlfunc/vec> for information on how to manipulate individual bits
1803 =head2 Integer Arithmetic
1805 By default, Perl assumes that it must do most of its arithmetic in
1806 floating point. But by saying
1810 you may tell the compiler that it's okay to use integer operations
1811 (if it feels like it) from here to the end of the enclosing BLOCK.
1812 An inner BLOCK may countermand this by saying
1816 which lasts until the end of that BLOCK. Note that this doesn't
1817 mean everything is only an integer, merely that Perl may use integer
1818 operations if it is so inclined. For example, even under C<use
1819 integer>, if you take the C<sqrt(2)>, you'll still get C<1.4142135623731>
1822 Used on numbers, the bitwise operators ("&", "|", "^", "~", "<<",
1823 and ">>") always produce integral results. (But see also L<Bitwise
1824 String Operators>.) However, C<use integer> still has meaning for
1825 them. By default, their results are interpreted as unsigned integers, but
1826 if C<use integer> is in effect, their results are interpreted
1827 as signed integers. For example, C<~0> usually evaluates to a large
1828 integral value. However, C<use integer; ~0> is C<-1> on twos-complement
1831 =head2 Floating-point Arithmetic
1833 While C<use integer> provides integer-only arithmetic, there is no
1834 analogous mechanism to provide automatic rounding or truncation to a
1835 certain number of decimal places. For rounding to a certain number
1836 of digits, sprintf() or printf() is usually the easiest route.
1839 Floating-point numbers are only approximations to what a mathematician
1840 would call real numbers. There are infinitely more reals than floats,
1841 so some corners must be cut. For example:
1843 printf "%.20g\n", 123456789123456789;
1844 # produces 123456789123456784
1846 Testing for exact equality of floating-point equality or inequality is
1847 not a good idea. Here's a (relatively expensive) work-around to compare
1848 whether two floating-point numbers are equal to a particular number of
1849 decimal places. See Knuth, volume II, for a more robust treatment of
1853 my ($X, $Y, $POINTS) = @_;
1855 $tX = sprintf("%.${POINTS}g", $X);
1856 $tY = sprintf("%.${POINTS}g", $Y);
1860 The POSIX module (part of the standard perl distribution) implements
1861 ceil(), floor(), and other mathematical and trigonometric functions.
1862 The Math::Complex module (part of the standard perl distribution)
1863 defines mathematical functions that work on both the reals and the
1864 imaginary numbers. Math::Complex not as efficient as POSIX, but
1865 POSIX can't work with complex numbers.
1867 Rounding in financial applications can have serious implications, and
1868 the rounding method used should be specified precisely. In these
1869 cases, it probably pays not to trust whichever system rounding is
1870 being used by Perl, but to instead implement the rounding function you
1873 =head2 Bigger Numbers
1875 The standard Math::BigInt and Math::BigFloat modules provide
1876 variable-precision arithmetic and overloaded operators, although
1877 they're currently pretty slow. At the cost of some space and
1878 considerable speed, they avoid the normal pitfalls associated with
1879 limited-precision representations.
1882 $x = Math::BigInt->new('123456789123456789');
1885 # prints +15241578780673678515622620750190521
1887 The non-standard modules SSLeay::BN and Math::Pari provide
1888 equivalent functionality (and much more) with a substantial
1889 performance savings.